Predictors of prognosis in acute myeloid leukemia : a clinical and epidemiological study

(1)

Department of Medicine, Division of Hematology

Karolinska University Hospital Solna and Karolinska Institutet, Stockholm, Sweden

PREDICTORS OF PROGNOSIS IN ACUTE MYELOID LEUKEMIA

A CLINICAL AND

EPIDEMIOLOGICAL STUDY

Åsa Rangert Derolf

Stockholm 2010

(2)

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet. Printed by Repro Print AB.

(3)

To my family and

in memory of my father

(4)

ABSTRACT

AML is a malignant disorder characterized by clonal expansion of immature myeloid hematopoietic stem cells, myeloblasts, in bone marrow, blood and/or other tissue. Despite advances in treatment the majority of patients eventually die from this aggressive disease.

We conducted a study including 9,729 AML patients diagnosed in Sweden 1973-2005 to define survival patterns over time. One-year relative survival ratios (RSRs) improved in all age groups. Improvement in 5-year RSRs was restricted to patients <80 years. The 5-year RSRs in the last calendar period were 0.65, 0.58, 0.36, 0.15, 0.05, and 0.01 for the age groups 0-18, 19-40, 41-60, 61-70, 71-80, and 80+ years, respectively. Intensification of induction and consolidation treatment, an increasing rate of allografted patients, a continuous improvement in supportive care measures, and a more precise risk stratification of patients are probably the most important factors contributing to the improvement. We also assessed the impact of socioeconomic status (SES) on survival in 9,165 patients with AML. Overall, higher white-collar workers had lower mortality compared to other SES groups (p=0.005). In AML patients, a consistently higher overall mortality was observed in blue-collar workers compared to higher white-collar workers in the last three calendar periods (hazard ratio [HR]=1.26; 95% confidence interval (CI) 1.05-1.51; HR=1.23; 1.05-1.45; HR=1.28; 1.04- 1.57, respectively). Differences in comorbidities, management, and life-style factors are likely to explain these findings.

We determined expression patterns of CD33 and CD15 in leukemic blasts from 129 patients with AML using flow cytometry (FC) and a standard panel of triple antibody combinations.

Five patterns, corresponding to the consecutive stages of myeloid differentiation, were identified [I:CD33-/CD15- (n=18), II: CD33+/CD15- (n=43), III: CD33++/CD15 heterogeneous (n=10), IV: CD33+/CD15+ (n=50), V: CD33-/CD15+ (n=8)]. Patients with pattern II had the highest relapse rate and shortest median overall survival (OS; 8 months), but they were also the oldest (median age 72 years) and had a high frequency of unfavorable cytogenetics. Pattern V patients had a short OS (median 14 months) even though they were the youngest (median age 50 years) and had high remission rate. Age (p=0.004), cytogenetics (p=0.011), CD15 expression (p=0.031), and the immunophenotypic classification (p=0.024) were all independent significant predictors for OS.

The presence of minimal residual disease (MRD) in AML patients in complete remission (CR) is a predictor of poor prognosis. We determined MRD status by FC in 45 AML patients

≤60 years old in first CR. MRD was determined after induction (MRD1; n=43) and/or at the end of post-remission chemotherapy (MRD2; n=31). Patients with detectable MRD at either time-point who underwent allogeneic or autologous stem cell transplantation (SCT) had significantly better 5-year relapse-free survival than patients not transplanted (MRD1: 83%, 54%, and 8%, respectively, p<0.0001; MRD2: 80%, 53%, and 0%, respectively, p=0.003).

We identified 11,039 patients with myeloproliferative neoplasms (MPNs) from the Swedish Cancer Registry and major hematology units. Through record-linkage with the Cancer Registry patients who developed AML (n=271) and myelodysplastic syndromes (MDS;

n=21) were identified. For each patient with a subsequent AML/MDS diagnosis (cases) two matched patients without AML/MDS (controls) were identified. After exclusions the final study population consisted of 162 cases (153 AML, 9 MDS) and 242 controls. 25% of patients with AML/MDS development were never exposed to cytotoxic agents. Compared to no hydroxyurea (HU) exposure the odds ratios (with 95% CIs) for 1-499 g, 500-999 g, >1000 g of HU were 1.22 (0.61-2.45), 1.41 (0.58-3.40), and 1.35 (0.55-3.32), respectively for AML/MDS development (not significant). In contrast, MPN patients who received radioactive phosphorus (P³²)>1000 MBq and alkylating agents >1 g had a 4.60-fold (2.15- 9.85; p<0.0001) and 3.39-fold (1.08-10.59; p=0.036) increased risk of AML/MDS, respectively. Lower exposures to P³²and alkylators were not associated with a significantly increased risk of AML/MDS.

(5)

LIST OF PUBLICATIONS

This thesis is based on the following papers, which are referred to in the text by their Roman numerals:

I. Rangert Derolf Å, Kristinsson SY, Andersson T M-L, Landgren O, Dickman PW, Björkholm M. Improved patient survival for acute myeloid leukemia: a population-based study of 9729 patients diagnosed in Sweden between 1973 and 2005. Blood 2009;113(16):3666-72.

II. Kristinsson SY, Rangert Derolf Å, Edgren G, Dickman PW, Björkholm M.

Socioeconomic differences in patient survival are increasing for acute myeloid leukemia and multiple myeloma in Sweden. J Clin Oncol 2009;27(12):2073- 80.

III. Rangert Derolf Å, Björklund E, Mazur J, Björkholm M, Porwit A.

Expression patterns of CD33 and CD15 predict outcome in patients with acute myeloid leukemia. Leuk Lymphoma 2008; 49(7):1279-91.

IV. Laane E, Rangert Derolf Å, Björklund E, Mazur J, Everaus H, Söderhäll S, Björkholm M, Porwit-MacDonald A. The effect of allogeneic stem cell transplantation on outcome in younger acute myeloid leukemia patients with minimal residual disease detected by flow cytometry at the end of post- remission chemotherapy. Haematologica 2006; 91(6):833-36.

V. Björkholm M, Rangert Derolf Å, Hultcrantz M, Kristinsson SY, Ekstrand C, Andreasson B, Birgegård G, Linder O, Malm C, Markevärn B, Nilsson L, Samuelsson J, Granath F, Landgren O. Treatment related risk factors for transformation to acute myeloid leukemia and myelodysplastic syndromes in chronic myeloproliferative neoplasms – a population-based nested case- control study. Manuscript to be submitted.

(6)

LIST OF ABBREVIATIONS

Allo-SCT AML APL

Allogeneic stem cell transplantation Acute myeloid leukemia

Acute promyelocytic leukemia ATRA

Auto-SCT CD CEBPA CI CR DA EBMT ET FAB FC FCM FLT3 G-CSF GM-CSF GVHD HR HU ICD ICE ITD JAK2 LGMS MDS MM MoAb MPN MRD NPM1 OR OS P³² PCR PMF PV RFS RSR SCT SES SIR SNOMED TRM WHO

All-trans retinoic acid

Autologous stem cell transplantation Cluster of differentiation

CCAT/enhancer binding protein alpha Confidence interval

Complete remission

Daunorubicin and cytosine arabinoside

European group for blood and marrow transplantation Essential thrombocythemia

French-American-British Flow cytometry

Flow cytometer

FMS-like tyrosine kinase 3

Granulocyte colony stimulating factor

Granulocyte macrophage colony stimulating factor Graft-versus-Host Disease

Hazard ratio Hydroxyurea

International Classification of Diseases

Idarubicin, cytosine arabinoside, and etoposide Internal tandem duplication

Janus kinase 2

Leukemia Group of Middle Sweden Myelodysplastic syndrome

Multiple myeloma Monoclonal antibodies Myeloproliferative neoplasm Minimal residual disease Nucleophosmin 1

Odds ratio Overall survival

Radioactive phosphorus Polymerase chain reaction Primary myelofibrosis Polycythemia vera Relapse free survival Relative survival ratio Stem cell transplantation Socioeconomic status Standardized incidence rate

Systemized Nomenclature of Medicine-Clinical Terms Treatment related mortality

World Health Organization

(9)

1 INTRODUCTION

1.1 HISTORY

In 1845 Rudolf Virchow (Figure 1) described a disease characterized by an enlarged spleen and excess of white blood cells. He named it ”leukemia” which is derived from Greek meaning ”white blood”.^1-3 At this time there had been other descriptions of patients with similar findings and it was generally considered that the white color of the blood was caused by the vessels being invaded by pus. John Hughes Bennett argued that this was the case and wrote a paper entitled “Two cases of disease and enlargement of the spleen in which death takes place from the presence of purulent matter in the blood”.¹ Alfred Donné, on the other hand, described several cases with a great excess of white blood cells in his book of 1844 and wrote that “Blood of such patients contains so many white blood cells that at first glance I thought they contained purulent matter.

In fact, I believe that the excess of white blood cells is due to an arrest of maturation of blood.”¹ However, Donné’s work was not acknowledged. Virchow continued his investigations and broke with the conventional wisdom and eventually recognized leukemia as an autonomous disease.

Nikolaus Friedrich was first to describe a case of acute leukemia in 1857. The clinical course differed from the two forms of indolent leukemia, splenic and lymphatic, described by Virchow.² At this time the function of the bone marrow was considered to be mechanical, to protect the blood vessels. Therefore it was not examined routinely at autopsy. In 1870 Ernst Neumann discovered changes in the bone marrow in a patient with leukemia and thus introduced the term “myelogenous”

leukemia. Another important contribution to the understanding of the leukemias was made by Friedrich Mosler in 1876, when he became the first physician to collect biopsy material from the bone marrow by sternal puncture. When Virchow developed his theories about leukemia, he believed that the white (or rather colorless) blood cells seen in leukemia were immature red cells. Paul Erlich was able to characterize the white blood cells further with his staining techniques and classified them into the lymphatic and myeloid systems in 1892.²

Despite early work of describing and classifying acute leukemias, it was not until 1976 that a uniform classification system was generally accepted. In 1976 the French-American-British (FAB) co-operative group published “Proposals for the classification of the acute leukaemias”⁴ in which they classified acute leukemias based on morphological characteristics of the leukemic blast in association with cytochemical reactivity patterns.

Six main types of acute myeloid leukemia (AML) were defined according to the direction of differentiation along

Figure 1. Rudolf Virchow (1821-1902)

(10)

one or more cell lineages and the degree of maturation. Modifications were made in 1985.⁵ The FAB classification was in use until 2001 when the World Health Organization (WHO) introduced a new classification in order to highlight the biologic and prognostic relevance of the cytogenetic abnormalities.⁶ It categorizes AML based on genetic findings, relation to cytotoxic therapy, and presence of myelodysplasia- related changes.⁷ Cases that do not fulfill criteria for inclusion in one of these groups are assigned to the group of “acute myeloid leukemia, not otherwise specified” and classified according to the major lineages involved and the degree of maturation. The WHO classification was revised in 2008^8-9 and is described in Table 1.

Tabell 1. Classification of acute myeloid leukemia according to the World Health Organization⁸ Acute myeloid leukemia with recurrent genetic abnormalities

t(8;21)(q22;q22); RUNX1-RUNX1T1*

inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11*

t(15;17)(q22;q12); PML-RARA* (acute promyelocytic leukemia) t(9;11)(p22;q23); MLLT3-MLL

t(6;9)(p23;q34); DEK-NUP214

inv(3)(q21q26.2) or t(3;3)(q21;q26.2); RPN1-EVI1

t(1;22)(p13;q13); RBM15-MKL1 (acute megakaryoblastic leukemia) Provisional entity: acute myeloid leukemia with mutated NPM1 Provisional entity: acute myeloid leukemia with mutated CEBPA

Acute myeloid leukemia with myelodysplasia-related changes Therapy-related myeloid neoplasms

Myeloid sarcoma

Acute myeloid leukemia, not otherwise specified**

Acute myeloid leukemia with minimal differentiation Acute myeloid leukemia without maturation

Acute myeloid leukemia with maturation Acute myelomonocytic leukemia

Acute monoblastic and monocytic leukemia Acute erythroid leukemia

Acute megakaryoblastic leukemia Acute basophilic leukemia

Acute panmyelosis with myelofibrosis Acute leukemia of ambiguous lineage Acute undifferentiated leukemia

Mixed phenotype acute leukemia with t(9;22)(q34;q11.2); BCR-ABL1 Mixed phenotype acute leukemia with t(v;11q23); MLL-rearranged Mixed phenotype acute leukemia, B/myeloid

Mixed phenotype acute leukemia, T/myeloid

Provisional entity: Natural killer (NK) cell lymphoblastic leukemia/lymphoma

*The diagnosis of AML is established without regard to blast cell count

**The category of AML not otherwise specified encompasses those cases that do not fulfill criteria for inclusion in one of the above described groups

(11)

1.2 ACUTE MYELOID LEUKEMIA 1.2.1 Definition

AML is a malignant disorder characterized by clonal expansion of immature myeloid hematopoietic stem cells, myeloblasts, in bone marrow, blood and/or other tissue. The diagnosis is typically established when at least 20% of nucleated cells in a bone marrow sample are myeloblasts (Figure 2). In AML with some specific genetic abnormalities the diagnosis is established irrespective of the blast cell count. Immunophenotypic analysis by flow cytometry (FC) has a central role in distinguishing between minimally differentiated AML and acute lymphoblastic leukemia (ALL). Mixed phenotype acute leukemia can contain separate blast populations of different lineages or one population with characteristics of different lineages. FC analysis is especially important when establishing the diagnosis in this leukemia subset.⁸

Figure 2. A bone marrow smear from a patient with acute myeloid leukemia showing a predominance of myeloblasts and presence of Auer rods

1.2.2 Epidemiology

Approximately 400 patients are diagnosed with AML in Sweden every year. This corresponds to an annual incidence of 3-4/100,000 inhabitants, which is similar to what is seen in other Western countries.^10-12 AML is diagnosed in all ages, but the incidence increases with increasing age and the median age at diagnoses is just below 70 years.

Males are slightly more affected than females.^10-12 1.2.3 Etiology and pathogenesis

The etiology of AML is unknown in most patients. However, some have a preceding diagnosis of another hematologic disease such as a myelodysplastic syndrome (MDS) or a myeloproliferative neoplasm (MPN). Another well established risk factor is

(12)

exposure to chemotherapeutic agents, especially alkylating agents and topoisomerase II inhibitors.¹³ Exposure to high doses of ionizing irradiation and chronic benzene exposure are other accepted risk factors.^{8, 14} Cigarette smoking has been reported to increase the risk for AML by 20-100%.^15-17

Development of AML in a patient is considered to be a process of multiple genetic changes in a hematopoietic stem cell. Mutations can either affect cell proliferation or cell survival (class I mutations) or they can affect differentiation and maturation of the hematopoietic cells (class II mutations). The presence of one of these mutation classes is not enough to induce leukemia, but when the two types are combined the disease can develop.^{13, 18-19}

1.2.4 Clinical signs and symptoms

AML is an aggressive disease where the symptoms typically appear rapidly. The expansion of myeloid blasts in the bone marrow leads to impairment of normal hematopoiesis. Thus, the patients often present with symptoms of anemia, bleeding, and infections.¹⁴ Skeletal pain is experienced by some patients and is directly related to the expansion of myeloblasts in the bone marrow. In addition, leukemic infiltration may produce organ specific symptoms. Hyperleukocytosis can occur and is associated with organ malperfusion and failure.^20-21 These patients are also at risk for cerebral and other hemorraghes.^21-22 Acute promyelocytic leukemia (APL; AML with t(15;17)(q22;q12);

PML-RARA) differs clinically from other AML subtypes in that patients carry a high risk of bleeding and thromboembolic events, even in the absence of leukocytosis, before and during early treatment.^23-25

1.2.5 Treatment

Intensive chemotherapy and achievement of complete remission (CR) is necessary for long-term survival in AML. Complete remission (CR) is defined as <5% leukemic blasts in the bone marrow in combination with a neutrophil count of >1.0x10⁹/L and a platelet count of >100x10⁹/L.²⁶ Curative treatment was not available until the late 1960’s, when the use of daunorubicin and cytosine arabinoside (DA) was introduced.^27-

28 These drugs combined made it possible to induce CR in AML patients. In a small subpopulation, there was even a potential of long-term survival. Improvement in supportive care over the years has enabled intensification of treatment. Several antileukemic drugs have been introduced since then, but no other drug combination has been convincingly shown to be better.^{14, 29-33} Thus, DA remains the cornerstone of AML treatment³⁴, though addition of etoposide³⁵ and the substitution of daunorubicin for idarubicin or mitoxantrone may improve overall survival in certain groups of patients.^36-38 In recent publications it is suggested that dose intensification of daunorubicin in both younger and older patients is tolerable and leads to superior survival.^39-40 The addition of granulocyte or granulocyte-macrophage colony- stimulating factor (G-CSF; GM-CSF) has been postulated to make leukemic blast more sensitive to chemotherapy. While patients belonging to the intermediate cytogenetic risk group may benefit from the addition of G-CSF⁴¹, no survival advantages have been shown regarding the addition of GM-CSF.^42-43

(13)

The concept of consolidation treatment, including high doses of cytosine arabinoside, has been developed to prevent relapse of the disease.^44-46 As a result, once the patient has achieved CR, consolidation treatment (usually 3-4 courses) is given. In addition, allogeneic stem cell transplantation (allo-SCT) was introduced as a clinical option in the mid 1970’s and reduces the risk of relapse and improves survival in selected patient groups.^47-51 Autologous SCT (auto-SCT) has been a part of AML treatment since the 1980’s but does not seem to significantly improve survival compared to conventional consolidation treatment.^{50, 52-53} An important difference between the two types of SCT is that the allo-SCT takes advantage of the donor’s immune system and there is clearly a graft-versus-leukemia effect. The drawback of the allo-SCT is the graft-versus-host disease (GVHD) which causes morbidity and mortality in a certain number of patients.⁵⁴ Allografted patients are also susceptible to bacterial, viral, and fungal infections several months after the transplantation has been performed, which is a major cause of therapy related mortality (TRM) in this group of patients.⁵⁵

APL is treated differently from all other subtypes of AML. The vitamin A derivative, all-trans retinoic acid (ATRA), introduced in the early 1990’s, has the ability to induce differentiation of leukemic promyelocytes in patients with APL and can induce CR as a single drug. In combination with an anthracycline-based chemotherapy the results are further improved.^{23, 56} In addition, arsenic trioxide has the ability to induce CR in patients with refractory and relapsed APL and is successfully used in combination with chemotherapy in this group of patients.^{23, 56}

When DA was first introduced, TRM was substantial because of a high incidence of severe infectious complications.⁵⁷ The introduction of empirical broad-spectrum antibiotics, antifungal and antiviral therapies, which are especially important in the allo- SCT setting, has reduced infection related mortality.^58-62 Altogether, in modern AML treatment, short-term mortality resulting from infectious complications has been reported to be as low as 4%.⁶³ In addition, patients treated with intensive chemotherapy and/or allo-SCT are often dependent on transfusion of erythrocytes and platelets for survival. Before platelet transfusions were made possible in the early 1960’s, many patients had fatal hemmorrhage.^{3, 64} Another important accomplishment in transfusion medicine is the testing for infectious diseases, especially in preventing transmission of

platelet transfusions empirical antibiotic prognostic significance molecular genetic treatment of cytogenetics markers

1960 1970 1980 1990 2000 2010 daunorubicin consolidation with

cytosine allo-SCT high dose arabinoside cytosine ATRA

arabinoside

A continuous improvement in antiviral, antifungal, and antibacterial treatment and transfusion medicine

Figure 3. Time-line showing important achievements in the management of acute myeloid leukemia

(14)

hepatitis and HIV.^64-65 In the allo-SCT setting, refinement of conditioning regimens and the use of larger amounts of stem cells have also contributed to lower TRM.^{59, 61} In a recent report from the Swedish Acute Leukemia Registry the overall allo-SCT TRM in patients treated 1997-2006 was estimated to be 14-19% in patients younger than 60 years of age.⁶⁶ Risk stratification based on cytogenetic analyses^67-69, and in recent years, molecular genetic analyses^70-73 has significantly contributed to an improved identification of patients who benefit from allo-SCT.

In all, survival of patients with AML is depending on successful induction treatment in combination with consolidation treatment and, in an increasing number of patients, allo-SCT. The importance of high quality supportive care during the entire treatment period cannot be overestimated.

1.2.6 Prognosis and prognostic factors

Despite advances in antileukemic treatment and supportive care the prognosis in AML remains rather poor. It is estimated that only 30-40% of younger patients and 10-20%

of older patients survive five years or more after diagnosis.^74-75 However, most survival data arrive from clinical trials that are associated with a various degree of patient selection and elderly/ frail patients are often not included.^76-77 Population-based data are scarce, though reports from England and USA suggest that survival has improved over the years.^{12, 78}

AML is a heterogeneous disease where outcome depends on many factors and treatment, including SCT, has to be tailored for each patient depending on the individual prognostic factors. The prognostic factors can be divided into three main categories 1) patient-related prognostic factors, 2) leukemia-related prognostic factors, and 3) response to treatment.

Patient-related prognostic factors

Old age is a well established independent predictor of poor prognosis in AML patients.51, 77, 79

AML in the elderly is also often associated with drug resistance and unfavorable cytogenetics.^79-80 Despite the general poor prognosis in elderly, some elderly patients seem to benefit from intensive chemotherapy and although few, long- term survivors exist.^81-82 Poor performance status and comorbidities are more common in elderly patients and often present in the same patient. However, they are also independently associated with increased induction treatment mortality.^{22, 83-84} For a number of cancers, most pronounced in cancer of breast, large bowel, bladder, and uterus, low socioeconomic status (SES) is associated with both increased risk and poorer outcome.⁸⁵ Shorter survival among young AML patients within lower SES groups has been reported⁸⁶ but there are few studies on the impact of SES in AML and an association between SES and outcome in AML has not been well established.^87-88 Leukemia-related prognostic factors

Chromosome abnormalities are detected in approximately 55% of adult AML patients.^89-90 In addition, more than 50% of patients with a normal karyotype have somatically acquired mutations in the nucleophosmin 1 (NPM1)⁷⁰, FMS-like tyrosine kinase 3 (FLT3)⁷¹, and/or CCAT/enhancer binding protein alpha (CEBPA) genes.^{72-73, 91}

(15)

Thus, about 75% of all AML patients can be genetically characterized. The information on cytogenetic and molecular genetic abnormalities is used to determine what risk category the patient belongs to (Table 2).

Table 2. Risk groups depending on cytogenetic and molecular genetic abnormalities³⁴ Favorable APL with t(15;17)(q22;q12): PML-RARA

inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11*

t(8;21)(q22;q22); RUNX1-RUNX1T1*

Mutated NPM1 without FLT3-ITD (normal karyotype) Mutated CEBPA (normal karyotype)

Intermediate Mutated NPM1 and FLT3-ITD (normal karyotype) Wild type NPM1 and FLT3-ITD (normal karyotype) Wild type NPM1 without FLT3-ITD (normal karyotype) t(9;11)(p22;q23); MLLT3-MLL

Cytogenetic abnormality not classified as favorable or unfavorable Unfavorable inv(3)(q21q26.2) or t(3;3)(q21q26.2); RPN1-EVI1

t(6;9)(p23;q34); DEK-NUP214 t(v:11)(v;q23); MLL rearranged.

del(5q) or -5 -7

abnl(17p)

complex karyotype**

* Exception: not favorable if kit-mutation of codon 816 is present

** Three or more chromosome abnormalities in the absence of one of the WHO designated recurring translocations or inversions, i.e. t(15;17), t(8;21), inv(16) el. t(16;16), t(9;11), t(v;11)(v;q23), t(6;9), inv(3)/t(3;3)

Therapy-related AML (secondary to treatment with chemotherapy and/or radiation) and AML secondary to a preceding hematological malignancy such as MDS or MPN are associated with lower rates of CR and a higher risk of relapse if CR is achieved.¹⁴ Also within this group of patients cytogenetic findings can be used for risk stratified therapy.⁹²

Patients with hyperleukocytosis at diagnosis have higher short-term mortality, mainly due to cerebral hemorrhage and/or respiratory failure, but once in CR their risk of relapse is probably not increased.^21-22

The use of immunophenotypic markers or patterns as prediction of prognosis in AML has not been widely accepted due to diverging results from the published studies.⁹³

Response to treatment

It is well accepted that a patient with a slow response to treatment, i.e. needing two or more cycles of chemotherapy to achieve CR, has a higher risk of relapse and shorter survival.⁹⁴ Results from several studies indicate that early evaluation of the bone marrow can be used to predict prognosis. The German cooperative AML-group has shown that the presence of more than 10% blast cells one week after end of induction chemotherapy is associated with lower rates of remission, shorter relapse-free and overall survival.^95-96 In addition, early clearance of peripheral blasts measured by flow cytometry has been reported to be associated with the achievement of CR.⁹⁷

(16)

Minimal residual disease (MRD) in AML is defined as remaining leukemic cells, usually detected by molecular genetic or flow cytometry methods, in a patient in morphological CR. The presence of MRD is clearly associated with a higher relapse rate and shorter survival.^98-99 For details see section 1.3.2.

1.2.7 Risk-adapted therapy

As previously mentioned, allo-SCT has the ability to prevent relapse in AML. The advantageous effect of allo-SCT is not only due to the high dose chemotherapy given as conditioning before the transplant. Rather, the graft-versus-leukemia effect is considered to be at least as important.⁵⁴ Allo-SCT with reduced intensity conditioning is an alternative for the older patients and the impact on outcome is currently evaluated in on-going studies.

Despite improvements in supportive care, TRM in allografted patients remains substantial. As a consequence, relapse-free survival may be improved, while no improvement regarding overall survival is seen. This challenging issue has led to the concept of risk-adapted therapy, i.e. patients with a low risk of relapse probably do not benefit from allo-SCT while patients with a high risk of relapse may.^47-49 The prognostic factors discussed above are used to select the strategy for the patients on an individual basis. Thus, patients allocated to the favorable risk group are generally not considered for allo-SCT, while patients with intermediate or high risk features are likely to benefit from allo-SCT.³⁴ In addition, the risk associated with the transplant itself as assessed by age, comorbidity and other transplant related risk factors needs to be taken into account when making individual clinical decisions.48, 100-101

1.3 FLOW CYTOMETRY IN ACUTE MYELOID LEUKEMIA

Flow cytometry (FC) is a laser-based technology which most widely used clinical application is to classify hematopoietic cells by cell surface immunofluorescence. The flow cytometer (FCM) scans single particles or cells as they flow in a liquid medium past an excitation light source. Analysis is based on size and granularity and whether the cells are carrying fluorescent molecules either in the form of antibodies or dies.

Photo detectors convert the signals to electric impulses, which are processed by a computer. Fluorescence occurs when a fluorescent molecule absorbs light at one wavelength, reaches an excited state and then returns to the ground state, emitting a light at a different (longer) wavelength.^102-103

Immunophenotypic characterization of blast cells by FC is used to distinguish between ALL and AML¹⁰⁴ and is also of great importance when establishing the diagnosis of mixed phenotype acute leukemia.⁸

Leukemic blasts in AML express normal myeloid differentiation antigens.¹⁰⁴ In addition, they often display leukemia-associated immunophenotypes (LAIPs; Figure 4), which are immunophenotypes that differ from their normal counterparts or is uncommon in regenerating bone marrow. LAIPs are usually divided into four groups based on the type of aberrant findings: 1) cross-lineage infidelity (i.e. lymphoid antigens expressed on myeloid blast cells), 2) asynchronous antigen expression (expression of a combination of myeloid-associated antigens which is not found in the

(17)

normal myeloid differentiation such as the co-expression of CD34 and CD15 or absence of CD13 in CD33 positive cells), 3) antigen over-expression (abnormally high expression of normal myeloid antigens), and 4) lack of antigen expression (absence of myeloid-specific antigens on myeloid blasts).^105-106 The leukemic clone may express more than one LAIP.

LAIPs may reflect underlying molecular abnormalities.^107-108 Specific immunophenotypic features have been found in most of the cytogenetically defined AML subsets.¹⁰⁸ Also, patients with defined molecular genetic markers often present with an associated immunphenotype.19, 109-110

Defining LAIP at diagnosis enables monitoring of MRD and certain immunophenotypic profiles have been proposed to be of prognostic relevance.

1.3.1 Immunophenotype in relation to prognosis

There are a number of reports on prognosis in relation to expression of individual antigens and combinations of antigens in AML. Some investigators have found an association between CD34 positivity and shorter survival.^111-112 However, other authors could not relate CD34 expression to outcome.^113-115 The current opinion is that CD34 expression can be associated with both favorable and adverse cytogenetics and thus, CD34 per se cannot be used to predict prognosis in AML.⁹³ CD56 expression has been associated with shorter survival in AML with t(8;21).¹¹⁶ Some^117-118, but not all¹¹⁹, studies showed similar results in other categories of AML. CD7 is the most common lymphoid marker observed in AML. Some authors have demonstrated shorter survival in patients with CD7+ AML, but it has not been confirmed by others (reviewed in⁹³).

CD15 expression has been associated with a higher CR rate after standard induction chemotherapy ^120-122 and with longer survival.¹²³

Combinations of antigens have been used with the aim to construct prognostic scores and new immunophenotypic classifications of AML. Casanovas et al.¹²⁴ formed a new classification based on seven antigens. In their study, overall survival was shorter in the group of patients having leukemic blasts expressing pan-myeloid markers and CD7. Another prognostic score was created by Legrand et al.¹²⁵, who demonstrated that no single antigen was of prognostic value, but that co-expression of pan-myeloid markers was associated with a better prognosis. Repp et al.¹²⁶ reported that certain

Figure 4. Examples of leukemia-associated aberrant phenotypes in acute myeloid leukemia a: Co-expression of CD34 and CD15 as an example of asynchronous antigen expression and,

b:Co-expression of CD56 and CD34 exemplifies cross-lineage infidelity

a b

(18)

individual antigens (such as CD9, CD13, CD33, and CD34) were related to a worse prognosis and if combined their prognostic discriminatory capacity improved.

Recently, it was suggested that the combination of CD33 and CD34 expression can be used for predicting prognosis in patients older than 60 years.¹²⁷ Thus, despite many efforts to define the prognostic value of immunophenotype in AML, this issue remains controversial.

1.3.2 Minimal residual disease detected by flow cytometry

MRD is defined as persistence of very low levels of leukemic cells in a bone marrow with morphological CR. The two most commonly used methods for investigating MRD are based on the detection, with high sensitivity, of either molecular or immunophenotypic markers expressed by the leukemic clone. The polymerase chain reaction (PCR) technique has a very high sensitivity (one target cell per 10⁴-10⁶ nucleated bone marrow cells) but the method is limited to the fraction of AML patients with specific genetic lesions.^128-129 FC allows a sensitivity of one leukemic cell per 10³- 10⁴normal bone marrow cells and can be used in more than 90% of AML patients.^130-

131 In MRD studies, follow-up samples can be acquired using a so-called live gate procedure when only cells with lineage associated markers are saved (CD19 in B-ALL, CD7 in T-ALL and myeloid marker of choice in AML). These cells are then screened for the possible persistence of residual cells with the same LAIPs as those identified at diagnosis.^{102, 132} The FC technique is under continuous development and five to eight color FCMs are becoming available, which will probably increase the sensitivity.¹³³ When molecular genetic techniques and FC are used to determine MRD in the same patient the concordance rate is high.¹³⁴

When FC is used to monitor MRD in an AML patient, the LAIP of the patient’s leukemic blasts has to be characterized at diagnosis. The most common measure points thereafter are at first CR and at the end of consolidation treatment. It is evident that the presence of MRD at these time-points is strongly associated with an increased risk of relapse 98-99, 106, 130-131, 135-136

and shorter overall survival. Some authors report a stronger association between risk of relapse and MRD-positivity post consolidation than after induction treatment.99, 130-131

However, it has also been suggested that MRD measurement at day 16 from start of induction treatment can be of prognostic relevance.¹³⁷ The presence of MRD seems to be associated with high risk karyotypes, but the prognostic significance probably remains within each karyotypic risk group.^98-

99, 134

There is no consensus on how to use the MRD information in the clinical setting.

The proposed use has been to select patients for SCT, though there are few reports in the literature regarding the benefit of SCT in MRD-positive patients. Auto-SCT does not seem to reduce the risk of relapse in MRD-positive patients¹³⁸, while there is some evidence that allo-SCT may improve relapse-free survival in these patients.^135-136 APL patients, on the other hand, are monitored regarding PML-RARA with PCR analysis and treatment is started if a molecular relapse is detected.^{23, 139} The use of FC to detect early relapse and start treatment before a clinical relapse is overt in non-APL AML has not been extensively investigated.

(19)

1.4 MYELOPROLIFERATIVE NEOPLASMS 1.4.1 Definition

Myeloproliferative neoplasms (MPNs) including polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF) are stem cell-derived clonal diseases characterized by proliferation of one or more of the myeloid lineages, i.e.

granulocytic, erythroid and megakaryocytic.¹⁴⁰ Initially, MPNs are characterized by a hypercellular bone marrow with an effective hematopoieses but the disease can progress into a stage with marrow failure due to myelofibrosis, ineffective hematopoieses or transformation into an acute blast phase.^141-143 The diagnostic criteria for PV, ET, and PMF according to the WHO classification of 2008 are given in Table 3.⁸ Chronic myelogenous leukemia is genetically characterized by the BCR-ABL1 fusion gene and considered a specific entity among the MPNs⁸ and will not be discussed further in this thesis.

Table 3. Diagnostic criteria for polycythemia vera, essential thrombocythemia, and primary myelofibrosis as defined by the World Health Organization classification of 2008⁸

Polycythemia vera: Diagnosis requires the presence of both major criteria and one minor criterion or the presence of the first major criterion together with two minor criteria

Major criteria

1. Hemoglobin >18.5 g/dL in men, >16.5 g/dL in women 2. Presence of JAK2 mutation

Minor criteria

1. Bone marrow biopsy showing hypercellularity for age with trilineage growth with prominent erythroid, granulocytic and megakaryocytic proliferation

2. Serum erythropoietin level below the reference range for normal 3. Endogenous erythroid colony formation in vitro

Essential thrombocythemia: Diagnosis requires meeting all four criteria 1. Sustained platelet count ≥450x10⁹/L

2. Bone marrow biopsy showing proliferation mainly of the megakaryocytic lineage with increased numbers of enlarged, mature megakaryocytes. No significant increase or left-shift of neutrophil granulopoiesis or erythropoieses

3. Not meeting WHO criteria for polycythemia vera, primary myelofibrosis, BCR-ABL1 positive chronic myelogenous leukemia, myelodysplastic syndrome, or other myeloid neoplasm

4. Demonstration of JAK2 mutation or other clonal marker, or in the absence of JAK2, no evidence for reactive thrombocytosis

Primary myelofibrosis: Diagnosis requires meeting all 3 major and 2 minor criteria Major criteria

1. Presence of megakaryocyte proliferation and atypia, usually accompanied by reticulin and/or collagen fibrosis, or in the absence of significant reticulin fibrosis, the megakaryocyte changes must be

accompanied by an increased bone marrow cellularity characterized by granulocytic proliferation and often decreased erythropoiesis

2. Not meeting WHO criteria for polycythemia vera, BCR-ABL1 positive chronic myelogenous leukemia, myelodysplastic syndrome, or other myeloid neoplasm

3. Demonstration of JAK2 mutation or other clonal marker, or in the absence of JAK2, no evidence that the bone marrow fibrosis or other changes are secondary to infection, chronic inflammatory condition, lymphoid neoplasm, metastatic malignancy, or toxic myelopathy

Minor criteria

1. Leukoerythroblastosis

2. Increase in serum lactate dehydrogenase level 3. Anemia

4. Splenomegaly

(20)

1.4.2 Epidemiology

MPNs are mainly disorders of the middle ages. The median age at diagnosis ranges between 60-70 years depending on MPN subtype, where ET is more common than PV and PMF in younger individuals.^144-146 Females are more commonly diagnosed with ET^146-147 while men are more frequent among PMF patients.^{146, 148} Some authors report a higher incidence of PV in men^149-150, but these findings have not been confirmed by others.^145-146 The annual incidence of all subtypes combined is estimated to be 6- 10/100,000 inhabitants.^{8, 144}

1.4.3 Etiology and pathogenesis

The underlying cause is unknown in most cases. A genetic predisposition has been reported in some families^151-152 and relatives of MPN patients seem to have a significantly increased risk of MPN.¹⁴⁶ The Janus kinase 2 (JAK2) mutation is present in 95% of PV patients and in approximately 50% of ET and PMF patients¹⁵³ and gives a proliferative advantage of hematopoietic precursor cells.¹⁵⁴ The mutation occurs at a primitive stem cell level and is chronologically an early event. However, there is evidence to suggest that the JAK2 mutation may not be the initial clonogenic event in PV or other MPNs and its presence may not be mandatory for endogenous colony formation.¹⁵³

1.4.4 Clinical signs and symptoms

Approximately half of all MPN patients are reported to be asymptomatic at diagnosis.^{140, 147} The disease is often indolent and the morbidity related to MPN is mainly related to venous or arterial thrombosis.^155-156 Hemorrhage may also occur, especially in ET patients.¹⁵⁷ Some patients experience pruritus (typically aquagenic and mainly in PV), erythromelalgia or other symptoms of acral ischemia. Splenomegaly is common and is often due to extramedullary hematopoiesis. In addition, some patients with PMF experience fatigue, weight loss, night sweats, and low-grade fever.

1.4.5 Treatment

The treatment in PV and ET is focused on reducing the risk of thromboembolic events and bleeding. Low-dose aspirin has been shown to reduce the risk of both arterial and venous thrombosis¹⁵⁸ and is recommended to almost all patients. Patients with PV are treated with phlebotomy to reduce the hematocrit below 0.45. In addition, patients who are considered at risk for thromboembolic events (i.e. patients above the age of 60 years, with platelet counts >1500x10⁹/L, and/or with a thromboembolic history) also receive cytoreductive treatment.¹⁵⁹ Cytoreduction can be accomplished by several different approaches. Hydroxyurea (HU) is the treatment of choice in patients above the age of 60. In the elderly patient, treatment with radioactive phosphorus (P³²) and alkylating agents is also an option. In younger patients, on the other hand, treatment with interferon is recommended as first line therapy. In addition, ET patients may be treated with anagrelide. Some PMF patients with anemia may be successfully treated with erythropoietin.¹⁶⁰ HU can be used to reduce spleen size and general symptoms.¹⁵⁹

(21)

1.4.6 Prognosis

The prognosis differs substantially between the subsets of MPNs. Life expectancy of patients with ET^{150, 161} has been reported to be similar to that of the general population while in PV patients life expectancy has been observed to be reduced.^{150, 162} PMF clearly has the worst prognosis with an average survival of less than five years.143, 148, 162 However, there is a well recognized risk of transformation to AML or MDS in a subset of patients in all three MPN subtypes. The risk has been reported to be highest in PMF followed by PV and ET (8-20%, 5-10%, and 2-5% at 5-15 years, respectively).^148,

161-167

Patients diagnosed with AML secondary to MPN have a very poor prognosis with a median survival of a few months.^{143, 163} In addition, patients with PV and ET have a risk of progression to PMF.^{141, 161} Risk factors at diagnosis of PV and ET that could be used to predict leukemic transformation are lacking. However, it is established that the risk of transformation is increased in patients treated with P³²or alkylating agents.165, 168-169

The issue of the leukemogenic potential of HU, on the other hand, remains controversial.¹⁷⁰ In fact, AML/MDS incidence rates of 10-14%^171-172 with HU used alone and 30% when preceded by busulphan treatment have been reported.¹⁷² Others have found no increased risk for transformation in patients treated with HU.^161,

167, 170, 173

The diverging results from the studies referred to above are likely explained by the fact that most of them are single institution studies including small numbers of patients with transformation to AML/MDS and, in many cases, a median follow-up time of less than ten years. The launching of large randomized trials addressing the issue of the risk for treatment related AML/MDS transformation has been hampered by the relative rarity of MPNs, late appearing transformation events in a long-term disease course and reluctance to randomize patients to potentially leukemogenic therapies.

1.5 SWEDISH POPULATION REGISTRIES

Sweden has a long history of population registries, the first was introduced as early as 1686 for military purposes, with the first report of survival in 1746. The personal identification code system for all Swedish citizens was established in 1947. Information regarding patients diagnosed with a malignant disorder is reported to the Swedish Cancer Registry which was established in 1958. Every physician and pathologist/cytologist is obliged by law to report each occurrence of cancer to the registry. The registry contains information on diagnosis, sex, date of birth, date of diagnosis, and hospital where the diagnosis was made.^{10, 174} In a recent validation study focusing on lymphoproliferative malignancies diagnosed 1964-2003 the completeness and overall diagnostic accuracy of the registry was found to be >90-95%.¹⁷⁵ In addition, there is the Swedish Adult Acute Leukemia Registry founded in 1997 by the Swedish Society of Hematology. This registry contains clinical information such as comorbidity, the patient’s cytogenetic risk group, applied treatment, and results of treatment.⁸¹ Allo- and auto-SCTs performed in Sweden are also reported to the European Group for Blood and Marrow Transplantation (EBMT) registry, which was established in 1974.

For each person the date and cause of death is registered in the national Cause of Death Registry. Statistics Sweden performed mandatory censuses every fifth year between 1960 and 1990 collecting information on individuals’ occupational status, income,

(22)

housing etc. This information is gathered in the Swedish National Census Database.¹⁷⁶ All these registries and the possibility of linkage between registries using the personal identification code system provide an excellent platform for performing epidemiological research.

(23)

2 AIMS

Overall aim

To improve management of patients with AML by identifying factors associated with risk of disease and outcome

Specific aims

To define outcome of patients with AML in Sweden over a long time period in relation to age, gender, year of diagnosis, and region of diagnosis and to relate the survival patterns to prevailing management strategies

To estimate the potential effect of socioeconomic status on survival in AML, using multiple myeloma as an indolent disease comparator

To seek for an association between the leukemic cell immunophenotype and outcome in AML

To evaluate the prognostic significance of minimal residual disease after induction and consolidation treatment with a special focus on the role of stem cell transplantation in younger adult AML patients

To define treatment related risk factors for transformation to AML in patients with myeloproliferative neoplasms

(24)

3 EPIDEMIOLOGICAL STUDIES ON SURVIVAL IN ACUTE MYELOID LEUKEMIA (I, II)

3.1 PATIENTS AND METHODS

In studies I and II we included information on all AML patients reported to the Swedish Cancer Registry from 1973 to 2005 using the International Classification of Diseases version 7 (ICD-7). The diagnoses coded as 2050 (acute myeloid leukemia), 2059 (acute myelomonocytic leukemia/acute myeloid leukemia non-specified), 2060 (acute monocytic leukemia), and 2069 (acute monocytic leukemia non-specified) were included. Using Systemized Nomenclature of Medicine-Clinical Terms (SNOMED) codes (introduced in 1993) we were able to define patients with acute promyelocytic leukemia (APL; SNOMED code 98663) diagnosed 1993-2005. Analysis of patients with a preceding myelodysplastic syndrome (MDS) was restricted to the same time period due to the fact that MDS was not reported to the Swedish Cancer Registry until the early 1990’s. Patients with APL or prior MDS were also included in the analysis of the whole AML cohort. Patients with a preceding cancer diagnosis including a hematological malignancy were included in study I but excluded in study II.

Additionally, in study II we included patients with multiple myeloma (MM) diagnosed from 1973 to 2005. Based on previous findings regarding the effect of hospital-type at diagnosis and gender on survival in MM¹⁷⁷, we were inclined to perform a study to assess the impact of socioeconomic status (SES) on survival in AML and MM. We chose to investigate these two hematological malignancies due to their different clinical characteristics. While AML is an aggressive malignancy which requires immediate management and is potentially curable¹⁴, MM is in most cases an indolent lymphoproliferative disorder with little or no prospect of cure.¹⁷⁸ We used occupational status as a proxy for SES, gathered from the Swedish National Census Databases¹⁷⁶, established from mandatory censuses conducted in 1960, 1970, 1980, and 1990. Seven SES groups were determined: higher white-collar worker, lower white-collar worker, self-employed, farmer, blue-collar worker, retired, and unknown. Information from the Cancer Registry included date of birth, sex, date of diagnosis, region, and hospital where the diagnosis was established. Date of death was obtained from the Cause of Death Registry. Information on the number of SCTs performed on AML patients during this time period was obtained from the EBMT registry (I).

3.1.1 Statistical methods

Relative survival ratios (RSRs) were computed as measures of AML survival (I).^179-180 An important advantage of using relative survival is that it does not rely on the accurate classification of cause of death. Instead, it provides a measure of total excess mortality associated with a diagnosis of AML irrespective of whether the excess mortality is directly or indirectly due to the cancer. The RSR is defined as the observed survival in the patient group divided by the expected survival of a comparable group from the general population, which is assumed to be free from the cancer in question. One-year, five-year, and ten-year RSRs can be interpreted as the proportion of AML patients who

(25)

survived their malignancy at one, five, and ten years, respectively. Expected survival was estimated using the Ederer II method¹⁸¹ from Swedish population life-tables stratified by age, gender, and calendar period. One-, five-, and ten-year RSRs were calculated for four calendar periods: 1973-1980, 1981-1988, 1989-1996, and 1997- 2005 and six age categories: 0-18, 19-40, 41-60, 61-70, 71-80, and older than 80 years.

In addition, patients diagnosed with APL or prior MDS were studied separately in two calendar periods (1993-1999 and 2000-2005). For APL patients, three-year RSRs were used as outcome variable due to reduced observation time. Poisson regression was used to model excess mortality.¹⁸² The estimates of this model are interpreted as excess mortality ratios. As an example, an excess mortality ratio of 1.5 for males/females indicates that males experience 50% higher excess mortality than females.

We estimated survival in relation to SES using the Kaplan-Meier method (II).¹⁸³ Secondly, the relative risk of death (any cause and cause-specific) in relation to SES and calendar period was estimated using Cox’s proportional hazards regression. We conducted both univariate and multivariate analysis, adjusted for sex, area of residence at diagnosis, age at diagnosis (≤54, 55-64, 65-72, 73-78, 79-83, or ≥84 years), and calendar period at diagnosis (1973-1979, 1980-1989, 1990-1999, and 2000-2005). To investigate whether mortality in relation to SES had changed over time, we also conducted analyses stratified by calendar period.

3.2 RESULTS AND DISCUSSION

3.2.1 Survival in acute myeloid leukemia patients (I)

A total of 9,729 AML patients (4,954 males and 4,775 females; median age 69 years) were included in the study. Forty percent of the patients were diagnosed in the Leukemia Group of Middle Sweden’s regions (LGMS; Stockholm-Gotland, Uppsala, and Örebro regions). A total of 949 SCTs were reported to the EBMT register during the study period, 626 allo- and 323 auto-SCT. More than half of the SCTs were carried out during the last calendar period with allo-SCT dominating.

We observed significant improvements in survival during the 33 year study period.

One-year RSRs improved significantly for all age categories (Table 4) while

Table 4. One-year relative survival ratios with 95% confidence intervals in acute myeloid leukemia

patients stratified by age category and calendar period Calendar period

Age category (years)

1973-1980 (95% CI)

1981-1988 (95% CI)

1989-1996 (95% CI)

1997-2005 (95% CI)

0-18 0.40

(0.30,0.49)

0.62 (0.51,0.70)

0.73 (0.63,0.81)

0.81 (0.73,0.86)

19-40 0.37

(0.31,0.44)

0.61 (0.54,0.68)

0.71 (0.64,0.77)

0.74 (0.67,0.80)

41-60 0.31

(0.27,0.36)

0.44 (0.40,0.49)

0.61 (0.57,0.66)

0.61 (0.56,0.64)

61-70 0.19

(0.16,0.23)

0.32 (0.28,0.36)

0.46 (0.41,0.50)

0.48 (0.44,0.52)

71-80 0.12

(0.08,0.14)

0.15 (0.13,0.18)

0.26 (0.23,0.29)

0.28 (0.25,0.30)

81+ 0.09

(0.05,0.13)

0.05 (0.03,0.09)

0.13 (0.10,0.17)

0.16 (0.13,0.19)

(26)

five-year RSRs improved for all but patients older than 80 years of age (Table 5). As expected, improvement in survival was most pronounced in younger patients.

However, even among patients up to 80 years of age diagnosed in the last calendar period a fraction of long-term survivors was observed. Ten-year RSRs differed only little from five-year RSRs, suggesting that most patients who survive for five years are cured from the disease.

Table 5. Five-year relative survival ratios with 95% confidence intervals in acute myeloid leukemia

patients stratified by age category and calendar period Calendar period

Age category (years)

1973-1980 (95% CI)

1981-1988 (95% CI)

1989-1996 (95% CI)

1997-2005 (95% CI) 0-18 0.17

(0.10,0.25)

0.31 (0.22,0.41)

0.53 (0.42,0.62)

0.65 (0.56,0.73)

19-40 0.09 (0.06,0.14)

0.21 (0.15,0.27)

0.38 (0.32,0.45)

0.58 (0.51,0.65)

41-60 0.06 (0.04,0.09)

0.12 (0.09,0.16)

0.24 (0.21,0.29)

0.36 (0.32,0.41)

61-70 0.04 (0.02,0.06)

0.07 (0.05,0.09)

0.14 (0.11,0.17)

0.15 (0.12,0.18)

71-80 0.03 (0.01,0.05)

0.02 (0.01,0.04)

0.06 (0.04,0.08)

0.05 (0.04,0.07)

81+ 0.03 (0.01,0.09)

0.00 (0.00,0.00)

0.01 (0.004,0.04)

0.01 (0.001,0.04)

Males had a 5% higher mortality compared to females during the first five years after diagnosis (p=0.032; Table 6), which is consistent with a previous report of female predominance among long-term survivors.⁷⁴ Overall, patients resident in LGMS regions had a 6% lower mortality during the first five years after diagnosis (p=0.006;

Table 6), mainly confined to differences observed in the first calendar period. There was no significant difference in excess mortality ratios between patients diagnosed at university compared to non-university hospitals.

One hundred and eleven patients with APL were diagnosed between 1993 and 2005, constituting 2.5% of all AML cases during this period. This reflects the lower incidence of this AML subtype in Scandinavia compared to that seen in Southern Europe, Latin America, and Asia.¹⁸⁴ Patients diagnosed with APL were younger (median age 54 years; 56 and 47 years in the two calendar periods under study, respectively) than those diagnosed with other subtypes of AML, which is consistent with the published literature.¹⁸⁴ The fact that APL is mainly a disease of younger patients may contribute to the differences in incidence of this disease between Scandinavia, where the population is older, and for example Latin America with a younger population. Overall three-year RSR in APL was 0.61, being 0.53 (95% CI:

0.38;0.66) in patients diagnosed in 1993-1999 and 0.69 (95% CI: 0.55;0.79) in patients diagnosed 2000-2005. Forty-eight (43%) of the APL patients had deceased at follow- up. One-month mortality rate was 23%; 27% in 1993-1999 and 18% in 2000-2005.

Bearing the limited number of APL patients in mind, short-term mortality appears lower in the later calendar period. An improved immediate management of this patient

Predictors of prognosis in acute myeloid leukemia : a clinical and epidemiological study

Department of Medicine, Division of Hematology

Karolinska University Hospital Solna and Karolinska Institutet, Stockholm, Sweden